AI Exposure Scores: what they measure, what they miss, and what comes next

The article discusses the introduction of GPTs are GPTs scores, which measure the share of occupational tasks that large language models can assist with, highlighting their methodological contributions and limitations. It identifies two significant gaps: the mismatch between static exposure scores and the dynamic policy questions they aim to inform, and the lack of coordination between researchers and policymakers in addressing these issues. The authors advocate for improved metrics and collaborative approaches to better inform policy decisions regarding the impact of AI on work.